Precision medicine in type 1 diabetes: comparing metabolic outcomes of Control‐ IQ and MiniMed 780G according to patient characteristics

Abstract Aims This study aimed to compare 12‐month metabolic outcomes in patients with type 1 diabetes (T1D) treated with either MiniMed 780G (Guardian 4) or Control‐IQ (Dexcom G6) automated insulin delivery (AID) systems and identify interaction with patient characteristics. Materials and Methods We conducted a single‐centre, retrospective study including all patients (aged ≥16) with T1D who were started on either MiniMed 780G or Control‐IQ between January 2021 and October 2022 and continued for ≥12 months. We used propensity score matching to compare the average marginal effects between MiniMed 780G and Control‐IQ regarding the primary outcome (time in range [TIR]) and secondary outcomes (time below range [TBR], glucose monitoring indicator [GMI] and coefficient of variation [CV]) after 12 months. We tested for interaction effects between baseline characteristics (age, sex, socio‐professional background, body mass index, insulin daily dose, carbohydrate counting practice) and treatment effect. Results We included 245 patients (58% women): 178 treated with Control‐IQ and 67 with MiniMed 780G. The mean ± SD age and haemoglobin A1c were 39 ± 15 years and 8.7 ± 1.8% (72 ± 20 mmol/mol) respectively. In the propensity score‐matched sample ( n  = 221), we observed significant differences in 12‐month TIR (MiniMed 780G minus Control‐IQ [95% CI]: 6.4 [3.4;9.5]), GMI (−0.42 [−0.59; −0.25]) and CV (−2.12 [−3.68; −0.55]), while TBR showed no significant difference (−0.04 [−0.47; +0.40]). The 12‐month TIR difference was consistent across subgroups, including baseline carbohydrate counting characteristics. Conclusion MiniMed 780G is associated with moderate metabolic superiority compared to Control‐IQ, without interaction with patient characteristics. These results suggest that neither model is more appropriate for certain populations, particularly patients without carbohydrate counting practice.